LOAD FORECASTING ANALYSIS SYSTEM FOR CALCULATING CUSTOMER BASELINE LOAD

• US 20100082172A1
• Filed: 09/25/2009
• Published: 04/01/2010
• Est. Priority Date: 09/25/2008
• Status: Active Grant

• Current Assignee: Korea Electric Power Corporation (Government of South Korea)
• Original Assignee: Korea Electric Power Corporation (Government of South Korea)
• Inventors: YU, In-Hyeob, KO, Jong-Min, JUNG, Nam-Joon, KIM, Young-Il
• Granted Patent: US 8,406,935 B2
• Time in Patent Office: Days
• US Class Current: 700/291
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 50/06 Energy or water supply
• CPC Class Codes: H02J 3/003 Load forecast, e.g. methods or systems for forecasting future load demand
• CPC Class Codes: Y02P 80/10 Efficient use of energy, e.g. using compressed air or pressurized fluid as energy carrier
• CPC Class Codes: Y04S 10/50 Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
• CPC Class Codes: Y04S 50/14 Marketing, i.e. market research and analysis, surveying, promotions, advertising, buyer profiling, customer management or rewards

Abstract

A load forecasting analysis system for calculating a customer baseline load (CBL). The load forecasting analysis system includes a CBL forecaster for receiving a load profile and providing a CBL forecasting method, a period selector for selecting conditions used to calculate the CBL using the load profile, a CBL processor for calculating a forecasting value by forecasting the CBL using the CBL forecasting method according to the load profile and the conditions, and a CBL determiner for calculating an error value by comparing the load profile with the forecasting value.

Claims

1. A load forecasting analysis system for calculating a customer baseline load (CBL), comprising:

a CBL forecaster for receiving a load profile and providing a CBL forecasting method;

a period selector for selecting conditions used to calculate the CBL using the load profile;

a CBL processor for calculating a forecasting value of the CBL using the CBL forecasting method according to the load profile and the conditions; and

a CBL determiner for calculating an error value by comparing the load profile with the forecasting value.

2. The load forecasting analysis system of claim 1, wherein the load profile is a 15-minute interval load profile converted into a relational database.

3. The load forecasting analysis system of claim 1, wherein the CBL forecasting method includes at least one of a method using an average load profile for n days (where n is 10, 5, or 3) immediately before a forecasting date and a method using an average load profile for 8 days excluding maximum and minimum load profile values out of the load profile for 10 days immediately before the forecasting date.

4. The load forecasting analysis system of claim 3, wherein the method using the average load profile for n days immediately before the forecasting date calculates a forecasting value F_t at a time t using equation

5. The load forecasting analysis system of claim 3, wherein the method using an average load profile for 8 days excluding maximum and minimum load profiles of the 10-day load profile immediately before the forecasting date calculates a forecasting value F_t at a time t using equation

6. The load forecasting analysis system of claim 1, wherein the conditions include at least one of an analysis object period for calculating the forecasting value, an event time period, and an incentive for calculating a reduction cost for reduced electric power use.

7. The load forecasting analysis system of claim 1, further comprising:

a short-term demand forecaster for receiving the load profile and providing a future value forecasting method of the load profile;

a forecasting processor for forecasting the amount of electric power use using various parameters and options used in the short-term demand forecaster; and

a forecasting determiner for forecasting the total amount of electric power use and the amount of short-term electric power consumption by class of industry and class of contract at intervals of 15 minutes according to a customer classification reference and selects a short-term forecasting model by comparing forecast models.

8. The load forecasting analysis system of claim 7, wherein the future value forecasting method includes at least one of a method using an average load profile for 10 days immediately before a forecasting date, a method using an average load profile for 8 days excluding maximum and minimum load profile values out of the load profile for 10 days immediately before a forecasting date, and a time series analysis method, and the time series analysis method includes at least one of a moving average method, an exponential smoothing method, and an autoregressive moving-average (ARMA) model.

9. The load forecasting analysis system of claim 8, wherein the moving average method calculates a future value F_t at a time t using equation

10. The load forecasting analysis system of claim 8, wherein the exponential smoothing method calculates a future value F_t at a time t using equation F_t=F_t-1+α

(D_t-1−

F_t-1) where F_t-1 denotes a future value at a time (t−

1), α

denotes a smoothing constant, and (D_t-1−

F_t-1) denotes a forecasting error at the time (t−

1).